Messages in this thread | | | Date | Thu, 03 Jun 1999 19:09:47 +0200 | From | Artur Skawina <> | Subject | Re: zero-copy TCP fileserving |
| |
Bjorn Wesen wrote: > > > no, Richard is right - on modern cpus the additional writes don't slow > > you down that much (eg on pii - only ~7% for fullsized eth frames).
> no CPU in the world has memory writes for free! i'll rephrase if its
from the cpus pov - some can (almost) do them for free; in certain cases. (write posting, caches etc). That does of course not mean the writes are completely free, it only means the slowdown is (much) less than 100%.
> the sort of thinking that goes "it's not noticeable on a $500 CPU right > now so we don't have to care" is what keeps M$ programs slow and intel > rich.
how about "it's not noticeable on a $50 CPU right now"? :^)
[ note that I cheated a bit - the stock kernel 686 routines lose ~30% while copying. Will post a patch soon. ]
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |